Chapter 2: Knowledge and Naturalism
by Dallas Willard
Dallas Willard's project is to argue that certain varieties of naturalism cannot accommodate knowledge or knowing in the world. Given the limited physical resources of a naturalist world, too many ingredients of knowledge are missing from such a world to account for our coming to possess knowledge. Willard states his thesis thus in his fourth paragraph:
"I will try to explain why narrower Naturalism or unqualified Physicalism cannot find a place for knowledge, and specifically for three of its essential components: truth, logical relations and noetic unity" (p.24).Notice that he qualifies naturalism with "narrower." He does this because naturalism has not historically been a unified position. Willard distinguishes "narrower" naturalism from "generous" naturalism in the essay. The former only permits physical entities into existence whereas the latter is not necessarily limited to only physical entities. Narrower naturalism, therefore, is stronger (or bolder) than generous naturalism. While Willard argues in the above quotation that narrower naturalism cannot find a place for knowledge, Willard argues elsewhere in the essay that generous naturalism also fails to accommodate knowledge because the position is imprecisely defined and winds up being vacuous. We will look at this last claim first.
Willard observes that generous naturalism is hard-pressed to specify itself concretely. If physical objects are not the only "natural" objects, then what else qualifies as "natural" and why? In an effort to solve this problem of specification, some generous naturalists have appealed to "the findings of science" to round out the term "natural." Willard writes:
"Indeed, many of the 'generous' naturalists of the mid-twentieth century gathered around Dewey and Sidney Hook identified naturalism precisely with acceptance of science and only science as the arbiter of truth and reality, and seemed, at least, to accept whatever came out of the pipe of 'scientific inquiry' as knowledge and reality" (p.30).Such an appeal to "scientific inquiry," however, generates problems for the naturalist. The sciences simply are not in the business of making sweeping claims about reality as a whole. Whatever the individual sciences have to say within the domains of their respective fields, they have nothing to say about domains outside of their fields:
"Methodological monism is an enduring aspect of generic naturalism, and modern naturalism is often specified simply in terms of an exclusive application of scientific method in all inquiries. But how can this method support claims about the nature of reality as a whole? For example, one might state that the only realities are atoms (quarks, strings, etc.) and derivatives thereof. But how is he to support his claim? It certainly cannot be derived from any specific science (physics, chemistry on up) or from any conjunction of specific sciences. And it is not to be derived through any application of experimental techniques within any science" (p.28).At best, a naturalist can appeal to the sciences to support claims to the effect that this or that exists, but such appeals to science will never amount to only this or only that exists 1. Willard goes on:
"[T]o suppose that a given scientific theory or conjunction of such theories provides an ontology constitutes a logical mistake, a misreading of what the theories say and imply. Those theories, and the bodies of knowledge wherein they are situated, actually say nothing whatsoever about the universe or about how it--the whole 'thing'--works" (p.28, original italics).In other words, if naturalism is to be a comprehensive philosophical worldview, and if it relies on the sciences to perform the task of specifying what the whole of reality contains, then it is in trouble because the sciences simply do not specify what the whole of reality contains. Whatever an individual scientist or acolyte of the sciences might say about the whole of reality, his or her account of the whole of reality was not derived from the sciences but rather added to the sciences. Thus, any variety of naturalism that depends on the findings of science for its ontology will be vacuous, for "science" simply says nothing whatever about the whole of reality.
Furthermore, there is the problem of specifying "successful science." Presumably, the naturalist who appeals to science to complete his ontology is not including failed scientific hypotheses in his ontology. For example, modern naturalists will not rank "phlogiston" up there next to oxygen on their list of existing entities, even though phlogiston was a properly scientific hypothesis in its time. Any present-day theory widely heralded as "scientific" may, for all that, still turn out to be false. The naturalist who weds his exposition of reality and knowledge to the findings of science is subject to oscillating ontological and epistemological facts. Insofar as the sciences are liable to find just about anything, then generous naturalism is liable to include just about anything. Positions that are so flexible as to say anything is the case end up saying nothing is the case. Willard again:
"[Specifying natualism in terms of 'successful science'] is vacuous in practice, for there is no way of identifying and accessing the 'successful science' which is proposed as defining naturalism. At most you get 'science now,' which is really only 'some scientist(s) now.' And certainly no science (including psychology) that was not naturalistic in some strongly physicalistic or at least empiricist sense would be accepted as 'successful' by those inclined to naturalism. Then we are back in a circle: naturalism in terms of science--but, of course, naturalistic science" (p.30).In my estimation, Willard is correct to assert that generous naturalism is vacuous--it does not make any definite claims about what does or does not exist. Since no one knows what science will "successfully" find in the future, and since generous naturalists are wedded to whatever is "successfully" found by science, no generous naturalist can specify a comprehensive ontology. As a result, the charge of vacuity appears to stick; such a vacuous position cannot have anything authoritative to say about knowledge or knowing, since it has nothing to say about anything.
So much, then, for generous naturalism. Willard next turns to "narrower" naturalism as the only palatable option left for the naturalist. This kind of naturalism, however, still stirs up severe philosophical difficulties:
"Naturalism has to be an honest metaphysics; and that metaphysics has to be 'unqualified physicalism'...But then a thinker who would be naturalist would feel pressure to have recourse to some specific a priori analyses to render his ontological specification of naturalism plausible....In addition to coming up with such a priori analyses, however, to turn to such inquiry as might produce them would be to break with the epistemological monism essential to naturalism and introduce something like a 'first philosophy.' This would be discontinuous with the empirical methods of the sciences. In showing it is right through a priori analysis, naturalism would simply give up the game" (p.30-1)."Unqualified physicalism" is certainly "honest" in that it makes a definite claim: The world contains physical entities and only physical entities. There is no subterfuge or ambiguity in that claim, no vacuous appeal to "the findings of science." The problem for the narrower naturalist, though, is that he or she is committed to empirical or sense-perceptible modes of justification, but the claim "the world contains physical entities and only physical entities" is not amenable to empirical justification. No laboratory experiment, for example, could verify that material entities and nothing else exist. By the naturalist's own epistemological standards, then, unqualified physicalism is not known to be true. It is in principle unverifiable by empirical methods. Unqualified physicalism is a metaphysical claim. Narrow naturalists subscribe to an epistemology that only deals in physical claims. As such, no amount of naturalist inquiry into the physical will yield metaphysical results. In order to make a metaphysical claim, one must enter the realm of a priori analysis, which is to say, one must use non-sense-perceptible modes of inquiry. This is traditionally known as "first philosophy"--a rational analysis of concepts, necessities, possibilities, and universals. But, this mode of analysis is in no way compatible with the naturalist's exclusive commitment to empirical methods. Thus, the instant a narrow naturalist turns to a priori analysis to justify his proposal of unqualified physicalism, he has, as Willard says, "given up the game."
Hence, naturalists are in a dilemma:
"In specifying what naturalism is, therefore, one seems to be in an inescapable dilemma. Either one must turn to a priori (non-empirical) analyses to establish its monism, which will refute naturalism's basic claim about knowledge and inquiry, or its claim will have to rest upon a vacuous appeal to 'science'" (p.31).Willard regards this dilemma, rightly, in my estimation, as a decisive refutation of naturalism as a philosophical thesis--that is, as a metaphysical proposition about all of reality. As such, whatever naturalism has to say about knowledge will be undermined. Willard, however, does not stop there. He goes on to explain in more detail why narrower naturalism cannot accommodate knowledge.
To begin his appraisal of narrower naturalism's compatibility with knowledge, Willard sketches what knowledge itself is. He identifies three essential components of knowledge--truth, logical relations, and noetic unity--and argues that narrower naturalism falls short of satisfying all three.
With respect to truth, Willard argues that truth is essentially a "matching up" or "correspondence" between representation and reality. For example, if one entertains the proposition "Earth is spherical," he or she is entertaining a representation--that is, he or she is producing a mental state with specific content and properties. Willard argues that this proposition is "true" if that representation accurately maps onto or matches up with reality--in this case, the shape of planet Earth. There are, then, two essential components of truth: representation and reality. Both must exist if truth is to exist.
Now, how does truth fit into the narrower naturalist picture? Willard argues that narrower naturalism fails to accommodate truth because, in a world of exclusively material entities, no representations of anything can exist:
"Suppose that we have an acceptable list of physical properties and relations....Let us agree that whatever goes on such a list will count as physical properties, and that narrow naturalism is the proposal to confine our inquiries and conclusions to whatever shows up on the list and combinations thereof....The argument, then, is simply that no such property or combination of properties constitutes a representation of anything" (p.39, original italics).Willard's list of physical properties might include items like "height," "weight," "mass," "spin," "charge," etc. As best I can tell, he is right to assert that no combination of these properties could obviously be taken to equal a representation of anything. If this is accurate, then narrower naturalism indeed fails to accommodate representational states, and therefore fails to accommodate truth. It then goes without saying that truth is an essential component of knowledge (absent truth, what would "knowledge" even mean?). Thus, narrower naturalism seems incapable of supporting knowledge. Willard concludes: "The ontological structure of knowledge cannot be present in the world of narrower naturalism" (p.40).
Having established that truth is absent from the narrower naturalist picture, Willard goes on to argue that logical relations also cannot exist in the naturalist world. Logical relations are essentially truth-preserving rules of inference. For example, if one knows that "a is true," one also knows that "not a is false." Much of what we ordinarily take to be knowledge depends on logical relations. But, if logical relations depend on the existence of truth (which they clearly do), and truth does not exist in the narrower naturalist world, then logical relations cannot exist in the narrower naturalist world. Thus, narrower naturalism fails to accommodate logical relations, and this amounts to another strike against narrower naturalism's ability to accommodate knowledge.
Finally, Willard argues that narrower naturalism fails to accommodate noetic unity. Noetic unity is basically an extensive network of interconnected mental states and properties that unite cohesively into an overall state of awareness. For example, the proposition "Smith is a bachelor" requires many interdependent pieces of semantic content. Among these pieces are marriage, male, human being, and wife, but also rules of English grammar (including subject, predicate, noun, verb, adjective, etc.), time, space, an external world, an interior world, and, most importantly, a system of logical relationship between all of them. If logical relations do not exist in the narrower naturalist world, and noetic unity depends on such relations, then noetic unity also fails to exist in the narrower naturalist world. Without the ability to synthesize or interconnect the various elements of a statement as basic as "Smith is a bachelor," narrower naturalism faces serious difficulties in accommodating knowledge. So, noetic unity appears to amount to strike three against narrower natualism's ability to support knowledge.
Willard concludes that knowledge is impossible in a world of narrower naturalism--that is, in a world where nothing but physical entities exist. He asserts that knowledge is clearly possible in our world--many people know many things--so our world must not be a narrower naturalist world. Combining this finding with the earlier finding that naturalism is in the dilemma of being either vacuous or self-defeating, we have extremely strong grounds for rejecting naturalism in any of its forms.
1. See also chapter 1 of this volume and my previous post for more on this topic.↩